Torque Specifications Explained: Why One Value Doesn’t Fit All Bolts
If you’ve ever asked “what’s the torque spec for an M12 bolt” and gotten three different answers from three different people, all of them were probably right — for the bolt they were picturing. An M12 property class 8.8 bolt, dry and black-oxide finished, an M12 10.9 bolt with a zinc-flake coating, and an M12 stainless A2-70 bolt with anti-seize on the threads all need meaningfully different torque values to reach the same clamping condition, even though “M12” is the only fact in the question. Torque is not a property of a bolt’s size. It’s the output of a small calculation that depends on grade, coating, lubrication, and what the joint is actually trying to achieve — and skipping that calculation is one of the most common, least visible causes of joint failure on site.
This article works through why that’s true, what actually goes into a correct torque figure, and where a single “spec sheet number” simply isn’t the right tool for the job.
Torque Isn’t the Goal — Clamp Load Is
The point of tightening a bolted joint isn’t to apply a specific number on a wrench. It’s to stretch the bolt elastically by a controlled amount, generating a clamping force — called preload — that holds the joint faces together with enough force to resist the working loads the joint will see in service: shear, vibration, thermal cycling, or a combination of all three. Torque is simply the easiest thing to measure at the point of installation that gives a reasonable, repeatable estimate of that clamping force. It’s a proxy, not the actual objective.
That distinction matters because the relationship between the torque you apply and the preload you actually achieve isn’t fixed — it’s governed by friction, and friction is the least consistent variable in the entire joint. The standard simplified formula engineers use to relate the two is:
T = K × D × F
Where T is torque, D is the nominal bolt diameter, F is the target clamp load (preload), and K is the “nut factor” or K-factor — a dimensionless number that bundles up thread friction and under-head friction into a single coefficient. Everything about “why one torque value doesn’t fit all bolts” comes down to K and F changing independently of D, even when D — the number stamped on the spec sheet as “M12” — stays the same.
The K-Factor: Why Friction Moves the Number More Than Size Does
K typically ranges from about 0.10 for a well-lubricated, coated joint up to 0.30 or higher for a dry, as-rolled or slightly corroded steel-on-steel joint — and that’s roughly a three-fold swing on identical bolts of the same diameter and grade, purely from surface condition. A few practical reference points:
- Dry, black-oxide or plain steel threads, uncoated and unlubricated: K commonly sits around 0.20–0.22.
- Zinc-plated or galvanized threads without added lubricant: often slightly higher than plain steel, around 0.20–0.24, because zinc can increase friction unless it carries a supplementary wax or lubricant topcoat.
- Lubricated or waxed threads (oil, anti-seize compound, or a lubricated coating): K commonly drops to around 0.12–0.16.
- Stainless steel threads, dry: often higher still, frequently 0.25–0.30+, because stainless has a strong tendency to gall (cold-weld) under friction — which is exactly why anti-seize compound is routinely specified for stainless fasteners, and why our stainless steel bolt, nut and washer range is best installed with a lubricant rather than dry.
Put two M12 bolts side by side — one dry black-oxide, one lubricated and zinc-plated — and the torque required for the same clamp load can differ by 30–40% purely from the K-factor difference. Neither number is “wrong.” They’re both correct for the specific surface condition they describe, which is precisely why a torque value copied from one job’s spec sheet onto a different job’s identical-looking bolts is a genuine risk, not a shortcut.
Property Class: The Second Variable Hiding Behind “M12”
Bolt diameter tells you nothing about the bolt’s strength grade, and strength grade is the other half of the torque equation because it sets the safe target preload (F). ISO 898-1 defines property classes for steel fasteners using a two-number code — 4.6, 8.8, 10.9, 12.9 — where the first number, multiplied by 100, gives the minimum tensile strength in megapascals, and the first number times the second (divided by 10) gives the approximate yield strength ratio. A class 8.8 bolt has a minimum tensile strength of 800 MPa and a yield strength around 640 MPa (80% of tensile); a class 10.9 bolt reaches 1,000 MPa tensile and roughly 900 MPa yield. That difference in yield strength directly raises the safe target preload for a 10.9 bolt over an 8.8 bolt of the same diameter — which is why upgrading a joint’s grade without recalculating torque either under-uses the stronger bolt’s capacity or, done carelessly in the other direction, overloads a lower-grade bolt with a higher-grade torque figure.
We’ve covered the IS, ISO, DIN and ASTM standards that define these property classes and their regional equivalents in more depth in our quick reference guide to fastener standards — worth reading alongside this article if you regularly cross-reference IS 1367 grades against ISO 898-1 classes or ASTM A325/A490 markings. And if you’re not sure how to identify which property class a bolt on your desk actually is from its head markings, our guide on reading a fastener specification sheet correctly covers exactly that.
Diameter Still Matters — Just Not Linearly
Diameter (D) sits directly in the torque formula, so it’s not irrelevant — a larger bolt of the same grade and friction condition does need proportionally more torque. But because bolt cross-sectional area (and therefore load-carrying capacity) increases with the square of the diameter while torque in the formula increases only linearly with D, torque requirements scale faster than diameter alone would suggest as bolts get larger. This is why halving a bolt’s diameter doesn’t halve its torque spec — it typically reduces it by considerably more than half, and why extrapolating a torque chart by “roughly doubling for the next size up” gets progressively less accurate at larger diameters. It’s also why pitch (coarse vs fine thread) introduces a further small correction that most simplified torque charts ignore entirely, which is one more reason a generic size-only chart is an approximation rather than a specification.
An Illustrative Example: One Diameter, Four Different Numbers
To make the “one value doesn’t fit all bolts” point concrete, here’s how a single M12 bolt’s approximate torque requirement shifts across four realistic combinations of grade, coating, and lubrication. These figures are illustrative, rounded, and based on commonly published engineering approximations for general guidance only — they are not a substitute for a torque value specified by a structural engineer, the fastener manufacturer’s data sheet, or your project’s own specification, particularly on structural, pressure-retaining, or safety-critical joints.
| Bolt Condition (M12) | Approximate K-factor | Illustrative Torque |
|---|---|---|
| Property class 8.8, dry, black oxide | ~0.20 | ~85 Nm |
| Property class 8.8, zinc-plated, lubricated | ~0.14 | ~62 Nm |
| Property class 10.9, dry, black oxide | ~0.20 | ~120 Nm |
| Stainless A2-70, dry, no anti-seize (not recommended) | ~0.28 | Unreliable — galling risk means torque no longer predicts preload consistently |
That last row deserves its own callout: beyond a certain friction level, particularly with dry stainless threads, torque stops being a reliable proxy for clamp load at all, because so much of the applied torque is being consumed fighting thread friction and incipient galling rather than stretching the bolt. This is one of several reasons stainless fasteners should always be installed with anti-seize or another approved lubricant — it isn’t just about ease of installation, it’s about keeping torque meaningful as a preload indicator in the first place.
To see where the ~85 Nm figure in the first row actually comes from, it helps to run the formula with real numbers rather than treat it as a black box. An M12 class 8.8 bolt has a tensile stress area of roughly 84 mm² and a minimum tensile strength of 800 MPa, giving a proof load in the region of 60–65 kN once a standard safety margin below full tensile capacity is applied as the target preload (F). Using K ≈ 0.20 and D = 0.012 m: T = 0.20 × 0.012 × 62,000 N ≈ 149 Nm at full proof load — but most general-purpose joints are deliberately targeted at a more conservative 60–70% of proof load rather than the maximum, which brings the working figure back down into the 85–100 Nm region shown in the table. This is also exactly why “torque to proof load” and “torque to a safe working preload” are two different numbers, and why a supplier’s recommended torque is usually noticeably below the absolute maximum the bolt could theoretically take.
Metric vs Imperial: Don’t Mix Unit Systems Either
Most of the fastener grades discussed above — ISO metric property classes, IS-standard bolts — are specified and torqued in newton-metres (Nm). But a meaningful share of imported machinery, older plant equipment, and some client specifications still use imperial fasteners (UNC/UNF threads, SAE grades) torqued in foot-pounds (ft-lb) or inch-pounds (in-lb). The two systems aren’t just a units conversion exercise — SAE grades (Grade 2, 5, 8) don’t map one-to-one onto ISO property classes (4.6, 8.8, 10.9), even though “Grade 8” and “class 10.9” are sometimes loosely treated as equivalent in casual conversation on site. Applying a torque figure calculated for an ISO 8.8 metric bolt to a superficially similar-diameter SAE Grade 5 imperial bolt — or converting Nm to ft-lb without checking whether the underlying grade assumption still holds — reintroduces the exact same “one value doesn’t fit all bolts” problem this article opened with, just via a different route. When a job mixes both systems, treat metric and imperial torque tables as two separate references, not one table with a conversion factor bolted on.
Where Torque Isn’t Actually the Control Method
For some joints, torque is deliberately not the primary way preload is verified, because the friction variability described above makes it too imprecise for the application. High Strength Friction Grip (HSFG) structural bolted connections — the kind used in steel structural steelwork where the joint relies on friction between clamped plates rather than the bolt shank bearing in shear — are the clearest example. These joints are typically tightened using a turn-of-nut method (a calibrated snug-tight position plus a specified additional partial turn) or verified using direct tension indicator (DTI) washers, which have small raised bumps that flatten measurably as the correct tension is reached — a physical, visual confirmation that doesn’t depend on assuming a K-factor at all. We cover HSFG joints specifically in our structural engineer’s guide to HSFG fasteners, and DTI washers in our DTI washers explained article — both are worth reading if your project involves friction-grip connections rather than standard bearing-type bolted joints.
It’s also worth knowing that torque sensitivity isn’t limited to heavy structural steelwork. Solar module mounting clamps, for instance, are torque-limited from the other direction — the concern is over-tightening cracking the panel glass or the cell beneath it, not under-tightening a structural connection, which is exactly why solar racking manufacturers publish tight torque ranges (not just minimums) for their clamp hardware. We go into this in our guide to hardware for solar mounting structures, where a “just tighten it firmly” instinct is actually the wrong instinct in the opposite direction from most structural bolting.
Tightening Sequence Matters As Much As the Number
A correct torque value applied in the wrong order can still leave a multi-bolt joint unevenly clamped, which is why bolt pattern and tightening sequence are specified alongside torque on flanges, base plates, and any joint with more than a handful of fasteners. Two practices account for most of this:
- Cross-pattern (star pattern) tightening, rather than working around the bolt circle sequentially, spreads clamping force evenly as it builds. Tightening bolts in sequential order around a flange instead of in a star pattern tends to cock the flange face slightly as each bolt is tightened, so the last few bolts end up compensating for a joint face that’s no longer sitting flat — leading to uneven real clamp load even though every bolt technically saw the same torque wrench setting.
- Multi-pass tightening, commonly a snug-tight pass (roughly 20–30% of final torque) followed by one or two further passes at increasing percentages of the target value, again in a star pattern, lets the joint faces bed down gradually and evenly rather than having one bolt fully torqued while its neighbours are still loose enough to let the flange shift.
This matters most on gasketed joints — pipe flanges being the clearest example — where an unevenly clamped gasket face is a leak risk even if every individual bolt shows the “correct” final torque reading. If your work involves flange bolting specifically, this sequencing detail is often more consequential to joint integrity than the torque figure itself.
Tools: Getting the Number You Calculated Actually Applied
A correct torque figure is only useful if the tool applying it is accurate. Three common torque tool types, in order of typical precision:
- Beam-type torque wrenches are simple, inexpensive, and reasonably accurate when read correctly (at eye level, at the point of maximum deflection), but depend entirely on the operator reading the scale correctly under load — awkward in tight or overhead positions.
- Click-type (micrometer) torque wrenches are set to a target value and give a tactile and audible “click” at that torque, making them faster and less operator-dependent for repetitive work, but they require periodic calibration since the internal mechanism can drift with heavy use.
- Digital/electronic torque wrenches display the applied torque directly and often log readings, which is valuable where torque records need to be retained for quality or audit purposes, but they’re more expensive and still require calibration on the same schedule as click-type wrenches.
Regardless of type, torque wrenches are precision instruments and should be calibrated on a recurring schedule — commonly every 12 months or after a set number of uses, whichever comes first, and always after any drop or impact. A torque wrench that’s drifted 15% out of calibration silently reintroduces exactly the kind of unpredictability this whole article is about avoiding.
For large-diameter structural and flange bolts — commonly M24 and above, where hand-torquing to the required value becomes physically impractical — torque multipliers (gearbox attachments that trade wrench speed for mechanical advantage) and hydraulic torque wrenches take over from manual click-type tools. Very large or highly critical joints sometimes move away from torque entirely in favour of hydraulic bolt tensioners, which stretch the bolt directly and measure elongation rather than inferring clamp load through a friction-dependent torque figure at all — the same underlying logic as DTI washers, applied at a much larger scale.
Common Torque Mistakes We See Repeated on Site
Reusing a torque figure from a similar-looking job without checking grade or coating. The single most common mistake — an M16 8.8 torque value applied to an M16 10.9 bolt because “it’s the same size,” under-tensioning a bolt that had more capacity to give.
Assuming lubrication doesn’t matter for a “quick job.” Skipping anti-seize on stainless threads to save time is one of the most common causes of galled threads and inconsistent, unreliable final torque on exactly the fasteners where friction variability matters most.
Using an uncalibrated or “by feel” approach on critical joints. Impact wrenches set by ear and hand-tightened-then-a-bit-more approaches have no reliable relationship to a calculated torque value at all, and shouldn’t be used on structural, pressure, or safety-critical connections.
Over-torquing to “be safe.” More torque isn’t automatically safer — beyond the bolt’s proof load, additional torque risks yielding the bolt, stripping threads, or, in coated fasteners, cracking the coating and exposing bare metal to corrosion (see our guide on preventing fastener corrosion in coastal and industrial environments for how a cracked coating shortens service life even when the joint itself doesn’t fail immediately).
Not re-checking torque after initial settling. Gasketed and multi-bolt joints commonly lose some tension in the first hours or days after installation as surfaces bed in — a spec that calls for a re-torque check at a defined interval exists precisely to catch this, and skipping it leaves the joint at whatever tension it settled to, not the one it was tightened to.
Frequently Asked Questions
Why can’t I just look up one torque value for a given bolt size? Because torque doesn’t depend on size alone — it depends on the target clamp load (which is set by the bolt’s property class) and the friction condition of the threads and bearing surface (which is set by plating, lubrication, and surface finish). Two bolts of identical diameter can have a correct torque that differs by 30% or more once grade and lubrication are accounted for, so a size-only chart is only ever a rough starting point.
What is a K-factor and why does it matter so much? The K-factor (or nut factor) is a number that represents the combined friction in the threads and under the bolt head or nut face, used in the formula T = K × D × F to relate applied torque to achieved clamp load. Because friction is affected by coating, lubrication, and surface condition, K can vary by two to three times between a dry, uncoated joint and a well-lubricated, coated one — which is the single biggest reason identical-looking bolts need different torque values.
Does a higher property class always mean a higher torque value? Generally yes, for the same diameter and friction condition, because a higher property class allows a higher safe target preload. But grade isn’t the only variable — a lower-grade bolt installed dry can sometimes require a similar or even higher torque number than a higher-grade bolt installed with effective lubrication, because the friction difference can outweigh the strength difference. This is exactly why grade alone, like diameter alone, isn’t enough information to set a torque value.
Is it true that stainless steel bolts need lower torque than standard steel bolts? Not necessarily lower — the key issue with stainless is that dry stainless threads have high and inconsistent friction, which makes torque an unreliable predictor of clamp load unless anti-seize or another lubricant is used. With proper lubrication, stainless bolts can be torqued according to their actual property class and diameter using standard formulas; without it, the achieved preload becomes unpredictable regardless of what torque figure is applied.
How often should torque wrenches be calibrated? A common baseline is every 12 months or after a defined number of uses, whichever comes first, with additional calibration checks after any drop, impact, or suspected damage. Torque wrenches used for safety-critical or audited work often follow a stricter interval set by the project’s quality plan.
Do direct tension indicator (DTI) washers replace the need for a torque value? For the joints they’re used on — typically HSFG and other friction-grip structural connections — yes, in the sense that DTI washers verify achieved clamp load directly by physical gap measurement rather than relying on torque as a proxy. They’re specified precisely because torque-based control is considered less reliable for these connection types, not because torque itself was calculated incorrectly.
Conclusion
Treat torque as the answer to a small calculation, not a lookup value, and most of the mistakes in this article stop happening on their own. Before you tighten anything: confirm the property class, confirm the coating and lubrication condition, and confirm which of those a torque figure someone hands you actually assumed — because a number without those three pieces of context is really just a guess with more confidence than it deserves. Get those three answers first, and the right torque follows from them rather than the other way around.
If your project involves a bolt grade, coating, or joint type you’re not confident torquing correctly, talk to our team about the specific fastener you’re using — we can point you to the manufacturer’s recommended torque range for the exact grade and finish you’ve ordered.